Before capacity expansion I would consider comparing demand versus our bottleneck, whereby the focus would be to improve that slowest bottleneck (by e.g. installing a quality assurance control point before the bottleneck).
The second option to consider would be to share the workload with another machine within the same plant, or with another plant.
The third option to consider would be to see whether the customer (demand) can accept deliveries in smaller batches, that fits within the output that we can realize.
The last option would be to indeed increase capacity. I think here it's a straightforward cost-benefit analysis, where you take things like distance to customer (for transportation cost if it's a new plant) etc.